To estimate the beta for a Brazilian telecommunications company, you have collected a sample of telecom peers in Latin America and Asia and peers in the United States and Europe. The median beta versus a world index is around 1.5 for the Latin American and Asian peers and around 0.9 for the U.S. and European peers (both subsamples are sufficiently large). What would you have to believe to choose either the Latin American/Asian peer beta or the U.S./European peer beta? What additional analyses would you undertake to test either choice?